Message-ID: <gh-pgjdbc-pgjdbc-3835-c3353274858@github.com>
From: "vlsi (@vlsi)" <noreply+vlsi@github.com>
To: "pgjdbc/pgjdbc" <noreply+pgjdbc-pgjdbc@github.com>
Date: Tue, 30 Sep 2025 18:06:56 +0000
Subject: Re: [pgjdbc/pgjdbc] issue #3835: Virtual Thread pinned in PostgreSQL JDBC Driver 42.7.8
In-Reply-To: <gh-pgjdbc-pgjdbc-3835@github.com>
References: <gh-pgjdbc-pgjdbc-3835@github.com>
List-Id: <gh-pgjdbc-pgjdbc.github.com>
X-GitHub-Author-Login: vlsi
X-GitHub-Comment-Id: 3353274858
X-GitHub-Comment-Type: issue_comment
X-GitHub-Issue: 3835
X-GitHub-Repo: pgjdbc/pgjdbc
X-GitHub-Type: comment
X-GitHub-Url: https://github.com/pgjdbc/pgjdbc/issues/3835#issuecomment-3353274858
Content-Type: text/plain; charset=utf-8

Just wondering: do you share the same connection across concurrent execute threads?

`QueryExecutorBase.getStandardConformingStrings` acquires a connection-scoped lock, so if you got pinning, then it looks like the same connection was used in a different thread which looks like an anti-pattern.

Do you have a reproducer for the issue?

In any case, it makes sense to move `createAction.create()` out of the synchronized block, however, I believe it will just mask the true issue of "sharing the same connection between concurrent execute threads" which you probably need to address in the app.